I have an offgrid inverter/charger - an MPPSolar PIP4048MS (4000W inverter, 48V system). By and large it has worked well for 7 years.
Recently I bought an MPPSolar PIP5048MS (5000W inverter, 48V system) as a replacement. However, it keeps on tripping the RCD in the distribution panel. (The radial circuits for the lights work fine, there is no problem).
This RCD controls 7 MCB circuits. Each MCB circuit is a ring circuit with maybe 7 or 8 sockets (outlets) on it.
When I connect up the inverter and turn it on (in battery mode, with no ac input connection), I then go to the distribution panel and turn on each MCB at a time in order not to overload the inverter. This is when the problem begins.
If there is no load on the MCB circuit the MCB lifts and the RCD remains 'up'. However, if there is a load on the circuit - such as a 'fridge, freezer, etc. the RCD trips. On one particular circuit there is a small pump that controls a reverse osmosis system. The pump only needs to be in standby (not actively running) to cause the RCD to trip. I have tried a variety of loads - toaster, 100W coffee grinder, electric drill. As soon as the load is applied the RCD trips. The only thing (that I have tried) that doesn't cause a trip is a small table lamp.
In both models of the inverter there is a neutral-ground bond in the inverter while the inverter is in battery mode - this is the only mode that I have tried the new inverter in.
Because of the time involved in exchanging the 2 inverters and vica versa (in order to test the PIP5048) I decided to set up a small replica of the distribution panel out in the shed. The same result - the RCD trips in the same circumstances as above.
At the behest of the MPPSolar support I removed the screw inside the inverter responsible for the neutral-ground bond. It made no difference. As a result they say that the problem isn't caused by earth leakage from inverter. The earth leakage must be coming from the loads. However, I put the following to them:
In the meanwhile can you answer this one question please? If there is leakage current coming from the laptop or 'fridge or toaster, etc., etc. then why doesn't the RCD drop right now when any one of these appliances is used (now that I am using the 'old' PIP4048 or if I use the EPEVER). Why does the RCD wait to trip until the PIP5048 is used?
They haven't answered the question - they just keep repeating that I need to measure the leakage coming from the loads. (By the way, I did use a 5kW EPEVER inverter/charger for a few weeks at the start of summer with no RCD trips). The RCD is a 63A 30mA DP type A RCD.
Where do I go from here?